List your company

Code Generation Security on Python Code Generation Security Dataset 75 cases

55Baseline Control RateGemini-3-Flash-Preview
36.2841.144650.86May 11, 2026

Evaluation results

MethodLinks
Gemini-3-Flash-Preview2026.055581.813.356.43294.44.1
GPT-5.2-chat2026.055485.210.757.430.798.11.4
GPT-5.1-chat2026.054985.79.361.225.391.75.4
DeepSeek-V3.22026.053783.38.261.119.297.21.4